Shape of my heart: Cardiac models through learned signed distance functions

Jan Verhülsdonk, Thomas Grandits, Francisco Sahli Costabal, Thomas Pinetz, Rolf Krause, Angelo Auricchio, Gundolf Haase, Simone Pezzuto, Alexander Effland
Proceedings of The 7nd International Conference on Medical Imaging with Deep Learning, PMLR 250:1584-1605, 2024.

Abstract

The efficient construction of anatomical models is one of the major challenges of patientspecific in-silico models of the human heart. Current methods frequently rely on linear statistical models, allowing no advanced topological changes, or requiring medical image segmentation followed by a meshing pipeline, which strongly depends on image resolution, quality, and modality. These approaches are therefore limited in their transferability to other imaging domains. In this work, the cardiac shape is reconstructed by means of threedimensional deep signed distance functions with Lipschitz regularity. For this purpose, the shapes of cardiac MRI reconstructions are learned to model the spatial relation of multiple chambers. We demonstrate that this approach is also capable of reconstructing anatomical models from partial data, such as point clouds from a single ventricle, or modalities different from the trained MRI, such as the electroanatomical mapping (EAM).

Cite this Paper

BibTeX
@InProceedings{pmlr-v250-verhulsdonk24a, title = {Shape of my heart: Cardiac models through learned signed distance functions}, author = {Verh\"ulsdonk, Jan and Grandits, Thomas and Costabal, Francisco Sahli and Pinetz, Thomas and Krause, Rolf and Auricchio, Angelo and Haase, Gundolf and Pezzuto, Simone and Effland, Alexander}, booktitle = {Proceedings of The 7nd International Conference on Medical Imaging with Deep Learning}, pages = {1584--1605}, year = {2024}, editor = {Burgos, Ninon and Petitjean, Caroline and Vakalopoulou, Maria and Christodoulidis, Stergios and Coupe, Pierrick and Delingette, Hervé and Lartizien, Carole and Mateus, Diana}, volume = {250}, series = {Proceedings of Machine Learning Research}, month = {03--05 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v250/main/assets/verhulsdonk24a/verhulsdonk24a.pdf}, url = {https://proceedings.mlr.press/v250/verhulsdonk24a.html}, abstract = {The efficient construction of anatomical models is one of the major challenges of patientspecific in-silico models of the human heart. Current methods frequently rely on linear statistical models, allowing no advanced topological changes, or requiring medical image segmentation followed by a meshing pipeline, which strongly depends on image resolution, quality, and modality. These approaches are therefore limited in their transferability to other imaging domains. In this work, the cardiac shape is reconstructed by means of threedimensional deep signed distance functions with Lipschitz regularity. For this purpose, the shapes of cardiac MRI reconstructions are learned to model the spatial relation of multiple chambers. We demonstrate that this approach is also capable of reconstructing anatomical models from partial data, such as point clouds from a single ventricle, or modalities different from the trained MRI, such as the electroanatomical mapping (EAM).} }
Endnote
%0 Conference Paper %T Shape of my heart: Cardiac models through learned signed distance functions %A Jan Verhülsdonk %A Thomas Grandits %A Francisco Sahli Costabal %A Thomas Pinetz %A Rolf Krause %A Angelo Auricchio %A Gundolf Haase %A Simone Pezzuto %A Alexander Effland %B Proceedings of The 7nd International Conference on Medical Imaging with Deep Learning %C Proceedings of Machine Learning Research %D 2024 %E Ninon Burgos %E Caroline Petitjean %E Maria Vakalopoulou %E Stergios Christodoulidis %E Pierrick Coupe %E Hervé Delingette %E Carole Lartizien %E Diana Mateus %F pmlr-v250-verhulsdonk24a %I PMLR %P 1584--1605 %U https://proceedings.mlr.press/v250/verhulsdonk24a.html %V 250 %X The efficient construction of anatomical models is one of the major challenges of patientspecific in-silico models of the human heart. Current methods frequently rely on linear statistical models, allowing no advanced topological changes, or requiring medical image segmentation followed by a meshing pipeline, which strongly depends on image resolution, quality, and modality. These approaches are therefore limited in their transferability to other imaging domains. In this work, the cardiac shape is reconstructed by means of threedimensional deep signed distance functions with Lipschitz regularity. For this purpose, the shapes of cardiac MRI reconstructions are learned to model the spatial relation of multiple chambers. We demonstrate that this approach is also capable of reconstructing anatomical models from partial data, such as point clouds from a single ventricle, or modalities different from the trained MRI, such as the electroanatomical mapping (EAM).
APA
Verhülsdonk, J., Grandits, T., Costabal, F.S., Pinetz, T., Krause, R., Auricchio, A., Haase, G., Pezzuto, S. & Effland, A.. (2024). Shape of my heart: Cardiac models through learned signed distance functions. Proceedings of The 7nd International Conference on Medical Imaging with Deep Learning, in Proceedings of Machine Learning Research 250:1584-1605 Available from https://proceedings.mlr.press/v250/verhulsdonk24a.html.

Related Material